Received for publication November 29, 1988and in revised form December 27, 1988
A Comparison between Quin-2 and Aequorin as Indicatorsof Cytoplasmic Calcium Levels in Higher Plant Cell
Protoplasts'
Simon Gilroy*, Will A. Hughes2, and Anthony J. Trewavas
Department of Botany, University of Edinburgh, Mayfield road, Edinburgh, Scotland, United Kingdom
ABSTRACT
Assessment of the regulation of plant metabolism by the cal-cium ion requires a knowledge of its intracellular levels anddynamics. Technical problems have prevented direct measure-ment of the concentration of intracellular Ca2+ in plant cells in allbut a few cases. In this study we show that electropermeabilizedprotoplasts of Daucus carota and Hordeum vulgare took up theCa2+ indicating fluorescent dye methoxyquinoline(O-aminophen-oxy)ethane-N,N,N',N'-tetraacetic acid (Quin-2) and the Ca2+ in-dicating photoprotein, aequorin. These protoplasts subsequentlyrecovered their plasma membrane integrity. However, up to 10%of intracellularly trapped Quin-2 was associated with a protoplastvacuolar fraction. Also, Quin-2 loading reduced total ATP levelsby approximately 60% and inhibited subsequent protoplast divi-sion whereas aequorin loading reduced ATP content by only 20%and did not prevent division. Therefore, the basal cytoplasmicCa2+ level measured with aequorin (less than 200 nanomolar)may more reliably reflect that found in vivo in the unperturbedprotoplast than that measured with Quin-2 (120-360 nanomolar).However, measurements made with aequorin were found to beinaccurate at Ca2+ levels below 200 nanomolar, Quin-2 provingcomplementary in indicating these low Ca2+ concentrations. Cy-tosolic Ca2+ was observed to increase on treatment with azideand silver ions.
Many higher plant processes and enzyme activities arethought to be regulated by changes in the level of cytoplasmicfree Ca2", [Ca2`]I (6, 12). However, the factors that regulateintracellular Ca2+ levels in plant cells are poorly understood.This has mainly arisen from technical difficulties in the directmeasurement of cytoplasmic Ca2+ levels; direct measurementhas generally been restricted to studies ofsingle cells amenableto indicator microinjection or impalement with Ca2+ selectivemicroelectrodes or to the very limited number of cell typesthat take up permeant esterified form of fluorescent Ca2+indicators (2, 4, 11, 14, 20, 21, 28). However, a potentially
' Supported by funds from the Agricultural and Food ResearchCouncil.
useful technique for loading plant protoplasts with Fura-2and Indo- 1 at low pH has recently been reported (4).Attempts at correlative biochemistry require procedures for
measuring cytosolic calcium synchronously in very largenumbers of cells or protoplasts. We recently showed (7, 8)that intracellular Ca2" levels could be measured in protoplastsuspensions from higher plant cells loaded with the fluorescentCa2+ indicator, Quin-2. The protoplasts took up Quin-2 afterthe plasma membrane had been reversibly permeabilized byelectroporation (7, 8). These protoplast suspensions are ame-nable to both calcium level estimations and biochemicalanalysis. However, loading animal cells with Quin-2 has beenreported to perturb cell metabolism and potentially disruptcellular Ca2+ levels (5, 24). Fluorescent indicators ofthe Quin-2 family have also been reported to redistribute from thecytosol to other compartments within the animal cell (24). Ifthis were to occur in the plant cell, it would interfere with thedirect estimation of free cytoplasmic calcium levels. Bush andJones (4) have reported the uptake of Fura-2 but not Indo-lby organelles of barley protoplasts. As well as directly meas-uring such cellular redistribution and degree of metabolicdisruption, the reliability of measurements made with thefluorescent Ca2t-indicators may also be assessed by compari-son with measurements made using an alternative [Ca2+]indicator; one chemically unlike Quin-2 and thus unlikely tobe subject to the same problems.Aequorin is a protein that emits light on binding Ca2+ (26).
The rate of light emission is proportional to the Ca2+ concen-tration and aequorin can be used to quantify Ca2+ levelsbetween 10-7 and 10-4 M. It has been used as a sensitiveindicator of cytoplasmic Ca2+ levels in animal systems (5, 26)and giant algal cells (22, 28). Aequorin is sufficiently differentfrom Quin-2 as to potentially cross check [Ca2+], measuredwith the fluorescent dye. Various forms of reversible permea-bilization have been successfully used to nondisruptively loadanimal cells with aequorin (5).We have previously used electropermeabilization to effect
the uptake of Quin-2 (7, 8), since Quin-2/AM did not per-meate the cell membrane in measureable quantities (7). Wecalibrated the size of electropores formed during this proce-dure and found they were large enough to admit dextrans ofup to 80 kD mol wt. Aequorin is a photoprotein of approxi-mately 20 kD and thus small enough to enter the plant cytosolduring electropermeabilization. It is also sufficiently differentto Quin-2 to provide a more rigorous assessment of Quin-2generated data. We report here a comparison of the usage and
482 www.plantphysiol.orgon December 26, 2018 - Published by Downloaded from
metabolic disruption induced by the application of Quin-2and aequorin to measure calcium levels in plant protoplasts.This is the first report of the use of aequorin for the measure-ment of cytoplasmic calcium levels in higher plants.
MATERIALS AND METHODS
Materials
Unless stated otherwise, reagents were of analytical gradeor higher and were obtained from either Sigma Chemical Co.(Poole, Dorset, U.K.) or BDH Chemical Co. (Poole, Dorset,U.K.).
Protoplast Preparation
Protoplasts of Daucus carota were prepared from suspen-sion cultured cells as previously described (8). Mesophyllprotoplasts of Hordeum vulgare were prepared as in Owen etal. (23).
Electroporation
D. carota protoplasts were electroporated and loaded withQuin-2 as previously described (8) with the following modi-fications for aequorin uptake. After isolation, protoplasts werewashed twice in electroporation buffer (250 mM sorbitol, 5mM Hepes, 1 mM EDTA, (pH 7.1) with 1 N KOH), resus-pended to 5 x 106 mL-' and incubated at 4°C for 10 min.Protoplasts were then permeabilized with a single 3 kV cm-',50 Ms (time constant), D.C. pulse, in the presence of 1 mgmL-' aequorin. The aequorin was supplied by either Dr. R.Blinks (Mayo Foundation) or Sigma. The aequorin from Dr.Blinks produced approximately 10 times the photons per mgthan the Sigma product. However, the Sigma aequorin provedadequate after an initial purification by HPLC (Gilson system,DuPont Zorbax GF250 column run in 5 mM EDTA, 50 mMKCI, 5 mM Hepes (pH 7.1) with 1 N KOH). Immediately afterpermeabilization, 1 mM MgSO4, 1 mM KH2PO4, 1 mM ATP,2% (w/v) sucrose were added (resealing buffer) and the pro-toplasts incubated at 4°C for 10 min. Protoplasts were thenresealed for 45 min at 26°C, sedimented (50g, 15 min) andresuspended to fresh medium with 500 Mm CaC12. Protoplastswere then washed free of remaining extracellular aequorin bysedimentation (200g, 25 min) through a 3/5/7% (w/v) sucrose
step gradient made up in resealing buffer with 500 Mm CaC12and without ATP; 108 protoplasts mL-' were assayed for lightemission by aequorin. H. vulgare mesophyll protoplasts wereelectroporated as above except that all media contained 500mM sorbitol and permeabilization was by a single 1.1 kVcm' , 100 Ms pulse. These conditions were optimal for theprotoplasts derived from the plant material grown under our
particular conditions. Protoplasts derived from different plantsources or plants and cells grown under different conditionsrequire modification and optimization of the permeabiliza-tion and resealing protocol. The field strengths (V cm-') usedfor the electroporation in this study selectively permeabilizethe protoplast plasma membrane while leaving organellarmembranes intact.
Indicator Calibration
The response of aequorin light emission and Quin-2 fluo-rescence to Ca2" concentration was calibrated both in situ inloaded protoplasts (internal calibration) and with the indicatorin a solution of an ionic composition thought to mimic theplant cytosol: 100 mM KCI, 20 mM NaCl, 1 mM KH2PO4, 1mM NaATP, 20 mM Hepes (pH 7.2), 1 mm free [Mg2"](external calibration). For internal calibration the protoplasts,in resealing buffer, 1 mm free [Mg2+J, were made Ca2+ perme-able with 50 gM ionomycin. The protoplasts were then ex-posed to media of known free Ca2" concentration set with 5mM EGTA and appropriate amounts of CaCl2 (calculated bya reiterative computer program, kindly supplied by the Uni-versity of Bristol Department of Biochemistry). IntracellularCa2+ concentration was then assumed to be buffered at theextracellular level set by CaEGTA. Ca2+ levels were related tothe rate of aequorin usage as detailed below and to theintensity of Quin-2 fluorescence as in Gilroy et al. (7) assum-ing a Kd of Quin-2 for Ca2+ of 115 nM.
Aequorin Calibration
Free calcium concentration can be estimated from the light(luminescence) emitted by aequorin (5). Emission ofa photonfrom aequorin by reaction with calcium, irreversibly dis-charges the molecule. In the presence of a very considerableexcess of aequorin, the free calcium level can be calculatedfrom the constant intensity of the emitted light. However, ifthe level of aequorin is relatively low then the emitted lightintensity (photons per minute) decays with approximatelyfirst order kinetics as the aequorin is irreversibly discharged.In this case a plot of logo0 light intensity (Li) against timeproduces approximately a straight line. The slope of the linevaries with the calcium concentration. High calcium producesa steep slope and low calcium a shallow slope. Due to therelatively small amounts of aequorin taken up by electroper-meabilized protoplasts, this 'rate of decay' method was em-ployed in this investigation.A further problem with the calibration of aequorin origi-
nates from both the variable uptake of aequorin into proto-plasts in different experiments and variable decay ofthe levelsof active indicator during an experiment. A normalizingprocedure was therefore employed (5). At the end of eachexperiment, 50 Mm ionomycin and 500 Mm CaCl2 were addedto the protoplasts to discharge any remaining aequorin. Thelight emission during this treatment was measured andsummed with the light emitted during the experiment fromthe point at which the Ca2' estimation was to be made, togive the total detectable photons present in the protoplasts atthat time, L,m. For calibration of the aequorin signal thenthe slope of logio (Li/L,,,a) was measured and compared witha standard curve.The relatively low light levels emitted from aequorin loaded
protoplasts were assayed with the sensitive photomultipliertubes of a Kontron scintillation counter (Intertechnique,France) with coincidence filter switched off. Light emissionwas continuously sampled, and averaged, every 28 s. Countswere corrected for 5% counter efficiency which was assessedfrom the photons detected from a known activity of luciferase
reacting with a known amount of luciferin under countingconditions identical to those of aequorin. The assay wasperformed using 1 x 10-8 units of luciferase in 1 mM MgSO4,0.6 mm ATP, 50 mm glycine (pH 7.7) and started with 0.1mM luciferin. Luciferin/luciferase and aequorin emit photonsof similar wavelength (26). Such a scintillation counter-basedsystem for the measurement of aequorin luminescence re-quires no specialized luminometer and could potentially beused by any investigator with a suitable scintillation counter.
Plant cells exhibit autoluminescence due, for example, tofree radical reactions during metabolism. This light emissioncoupled with a 'dark current' from the measurement appara-tus provides a background to the aequorin luminescenceestimation and was subtracted from measurements prior tocalculation of log10 (Li/Lmax). Background luminescence rep-resented less than 30% of the signal from aequorin loadedprotoplasts.
Permeabilization, Fluorimetric and ATP Assays
The measurement ofprotoplast permeabilization and fluor-imetric assays for Quin-2 were performed as in Gilroy et al.(7). ATP content was assayed by the firefly luminescencemethod. Each assay (1 mL) contained: 108 protoplasts, 5 mmMgSO4, 1.5 mM NaN3, 0.5 mM EDTA, 50 mM Tris-HCl (pH7.0), 1 mg BSA, and 4 mg luciferin/luciferase reagent (Sigma,Poole, Dorset, U.K.). The light from the luciferin/luciferasereagent was at least 50-fold greater than the signal fromprotoplasts loaded with aequorin and was measured in a BairdNova spectrofluorimeter configured for bioluminescencemeasurement. However, the aequorin present in the loadingincubation (1 mg mL-') was found to interfere with theestimation of light emission in the luciferin/luciferase reac-tion. Therefore, Ca2"-discharged (nonluminescent) aequorinwas used to load protoplasts for these assays. Prior to use, theaequorin was incubated with 100 /uM CaCl2 for 30 min afterwhich period no Ca2' dependent light emission could bedetected. This Ca2+-discharged aequorin was then loaded onto a G25 sephadex column preequilibrated in 250 mm sorbi-tol, 5 mm EDTA, 50 mm KCI, 5 mM Hepes (pH 7.2) with 1N KOH, and centrifuged at 3000g for 10 min to exchange theCa" containing buffer for the buffer described above. Also,ATP was omitted from the resealing cocktail (1 mM MgSO4,1 mm KH2PO4, 1 mm ATP, 2% [w/v] sucrose) added afterthe electroporation to aid membrane resealing, as exogenousATP would have interfered with the measurements of endog-enous protoplast ATP levels.
Vacuole Isolation
D. carota protoplasts were fractionated essentially as inKringstad et al. (15). Protoplasts were prepared as in Gilroyet al. (8) in media supplemented with 500 mm sorbitol. Afterisolation, protoplasts were resuspended to 2 x 106 mL-' in:250 mm sorbitol, 5 mm EGTA, 25 mM Tris HCI (pH 8.0),and stirred at 4°C (30 rpm, magnetic stirrer). The lysedprotoplasts were passed through a 20 uM nylon mesh and thefiltrate spun at 3000g for 120 min at 4°C through a 2.5/5/7.5% (w/v) ficoll gradient made up in 250 mM sorbitol, 25mM Tris-HCl (pH 8.0). Vacuoles were collected at the 2.5/
5% ficoll interface. Organelles, unlysed protoplasts, and debriscollected in the pellet. Vacuoles were visualized by fluorescentlabeling. Prior to protoplast formation and vacuole isolationcells were incubated for 16 h in 0.5% (w/v) 5-carboxyfluores-cein (5-CF) which was taken up by the vacuoles (13). 5-CFloaded vacuoles were visualized with conditions as for fluo-rescein (19). To assess the contamination of the vacuolar,supernatant, and pelleted fractions by other cellular fractionsmarkers for various cellular compartments were assayed.These were: cytosol, PEP carboxylase assayed as in Wong andDavies (29); mitochondria (taken as a typical organelle), fu-marase assayed as in Hall and Moore (10); vacuoles, 5-CFfluorescence (excitation 460 nm, emission 550 nm) measuredin a Perkin-Elmer LSSB spectrofluorimeter. Although thevacuolar specificity of 5-CF has not been proven, it wasaccumulated by vacuoles of D. carota (13) and provided aqualitative indication of vacuolar distribution. The vacuolarmarker enzymes acid phosphatase and a-mannosidase (10)could not be reliably detected in the protoplasts used in thisstudy.
[3H]Quin-2 Production
[3H]Quin-2 was obtained by incubation of 1 ,uCi 'H-Quin-2/AM (Amersham International plc., Bucks., U.K., 6.9 mCi/mMol) with 2 x 106 cells of D. carota for 60 min in electro-poration buffer. The cells were then sedimented, 5OOg for 10min and the supernatant filtered through 20 ,um nylon meshand 0.2 ,um nitrocellulose (Millipore, U.K.). The filtrate wasused for [3H]Quin-2 uptake experiments. Incubation with thecells had completely hydrolyzed Quin-2/AM to Quin-2 in themedium, as assessed by the change in fluorescence emissionspectrum (excitation, 340 nm) from that of Quin-2/AM tothat of free Quin-2 (27). Also, the product of Quin-2/AMhydrolysis comigrated with pure Quin-2 on TLC plates (F254TLC plates, Merck, FRG) solvent system 9:1 chloro-form:methanol. The TLC spots were visualized by UV illu-mination (UVS-54 UV lamp; Ultraviolet Products Inc., SanGabriel, CA).
[3H]Quin-2 UptakeUptake of [3H]Quin-2 by isolated vacuoles was performed
on a vacuolar fraction prepared as above. Two mL ofvacuoleswere incubated with 22,000 cpm of [3H]Quin-2 and at inter-vals samples were removed, washed three times on WhatmanGF-C filters with 250 mM sorbitol, 1 mM Quin-2, 1 mM CaCl2,Tris-HCl (pH 8.0) and the filter counted by liquid scintillation(2:1 Toluene: Triton X-100 with 0.7% (w/v) butyl-PBD).Uptake was corrected for 56% breakage ofvacuoles as assessedby the ratio of the fluorescence of the vacuolar fraction from5CF loaded protoplasts (138 fluorescence units) to that in thefilter washings (78 units).
RESULTS
Aequorin Uptake into the Cytosol
On electropermeabilization in aequorin, 108 protoplasts ofD. carota took up aequorin equivalent to 7 to 8 x 10'0
detectable photons. During the resealing incubation prior toCa2+ measurements, this luminescence decayed by up to 80%.The remaining luminescence remained associated with theloaded, resealed protoplasts despite rigorous washing in mediacontaining 500 ,M CaCl2, which would have discharged anyextracellular aequorin. However, adding ionomycin, an ion-ophore highly selective for Ca2l (17), caused a burst of lightemission (Fig. IA). This suggests that aequorin was taken upwhen the plasma membrane was electropermeabilized andtrapped intracellularly when the electropores resealed. The
41)
cDC.)
E-j
3.4-
c 3.20
X 3.0
a 2.84-.-C' 2.60
o 2.4
0C; 2.2
2.0
C.0 1.0
- 0.8e.1-:0.6
° 0.4c)0
Cl)cn
-0.2
o3
0
0 10 20 30Time (minutes)
0 5Time
1 0
(minutes)1 5
Figure 1. Light emission above protoplast autoluminescence of D.carota protoplasts loaded with aequorin (A) and fluorescence abovebackground of protoplasts loaded with Quin-2 (B). Protoplasts weretreated with 10 gM ionomycin added at the arrow. Extracellular Ca2+concentration was 500 jM. Loading of Quin-2 and aequorin wasperformed as in "Materials and Methods."
intracellular aequorin would therefore be protected from theCa2" in the medium by the protoplast plasma membrane untilthe ionophore was added. Note that the light intensity quicklyrises but then falls even though the elevated cytosolic calciumlevel should be maintained by the presence of the ionophore.These kinetics are observed because the elevation of cytosoliccalcium discharges the limited amount of aequorin taken up.High levels of calcium giving rise to high rates of decay inluminescence.
Figure lB shows the response of the Quin-2 fluorescencewhen ionomycin was added to equivalent, but Quin-2 loaded,protoplasts. Thus, both aequorin and Quin-2 permitted thequalitative detection of fairly rapid changes in calcium level.However, the requirement in this study of measuring a log,0decay rate for aequorin over several minutes to quantify theCa2+ level (see below) made this indicator less suitable, underthe measurement conditions we have used, than Quin-2 forthe quantitative assay offast transients in cytoplasmic calciumlevel. The rapidity of response and elevation of indicatorsignal on the application of ionomycin suggests that bothindicators were loaded into a cellular compartment which hasboth a rapid exchange with the medium and a low restingCa2+ level, probably the cytosol (see below).
Directly after loading, protoplasts showed a high rate ofaequorin luminescence which decayed to a steady level over10 to 15 min (Fig. 1A). However, active aequorin was presentwhen the steady luminescence level was finally attained asapplication of ionomycin caused an intense burst of Ca2+dependent luminescence (Fig. IA). Therefore, the electropor-ation loading method may have initially disturbed cellularCa2+ metabolism causing cytoplasmic calcium levels to riseand so increase aequorin light emission. The subsequentdecline in luminescence would then represent the recovery ofa low, unperturbed cytoplasmic Ca2+ level. An initial, 50 min,elevation of intracellular calcium was noted by Williamsonand Ashley (28) upon microinjection of aequorin into Characells and attributed to a similar cause.
In order to confirm the presence of the trapped aequorinin the protoplast cytosol silver ions (1 mm final concentration)were added to the protoplast medium. Silver ions are a potentinhibitor of aequorin light emission causing its nonlumines-cent discharge (26). Addition of silver ions should poison anyaequorin adhering to the extracellular face of the protoplastwithout parallel emission oflight. However, as Figure 2 shows,on addition of 1 mM Ag+ to aequorin loaded protoplasts asmall but definite spike in luminescence was observed, fol-lowed after several minutes by a decline. The amplitude ofthe initial spike is probably underestimated due to the aver-aging of the photon emission by the scintillation counter usedfor the assay. Nevertheless, this must originate from an ele-vation of cytoplasmic calcium resulting from the addition ofthe silver ions to the protoplasts. The decline in luminescencewhich follows may result from this elevated calcium leveldischarging a significant proportion of the intracellular ae-quorin. However, penetration of the plasma membrane bysilver ions with the non-luminescent discharge of aequorincould also contribute. Therefore, a reliable estimation of theelevation in cytoplasmic calcium level induced by the appli-cation of silver ions could not be made. Silver ions interfered
Figure 2. Effect of silver ions on the light emission, above back-ground, of D. carota protoplasts loaded with aequorin. Protoplasts ofD. carota were loaded with aequorin by electroporation. One mmAgNO3 was added at the arrow. The extracellular calcium concentra-tion was 500 !M. The maximum amplitude of the transient rise in lightemission after application of Ag+ may be underestimated due to theaveraging of the luminescence signal by the scintillation counter usedfor the luminescence assay. Note log10 scale for luminescence.
with Quin-2 fluorescence (data not shown). Thus, parallelexperiments to quantitate the effect of silver on cytoplasmiccalcium levels using Quin-2 could not be performed. Silverhas been reported to inhibit the activity of a Ca2+-ATPase(Ca2+ pump) in animal cells (9). Given the widespread use ofsilver ions to antagonise the effects of ethylene (16) an un-
ambiguous measurement of the effect of this ion on cyto-plasmic calcium levels should be of some interest.
Metabolic Disruption Induced by Indicator Uptake
To be useful in the measurement ofcalcium concentrationsin the cytoplasm, Quin-2 and aequorin must not perturb thecellular Ca2"-regulatory system. Therefore, we measured theeffects of loading these indicators on protoplast viability,membrane integrity, and ATP levels. The latter measurementgives a limited but indicative pointer to metabolic status.
After uptake of Quin-2 or aequorin, protoplast plasmamembrane integrity was seen to recover. Approximately 80%of the protoplasts were permeabilized directly after electro-poration reducing to 32% (Quin-2) or 24% (aequorin) after 1
h resealing incubation, as is shown in Figure 3. This compareswith 14% rising to 18% permeabilized after an equivalent 1
h incubation in unelectroporated controls. Osmotic activity,the ability to swell and shrink on changing the osmoticstrength of the medium was also recovered (data not shown).Similarly, after the 1 h resealing incubation, viability, as
assessed by fluorescein diacetate (FDA) staining, was 76 +
4% n = 30 (Quin-2 loaded), 70 ± 9% n = 8 (aequorin loaded)compared with 83 ± 11% n = 30 for unelectroporated con-
100
'aVU1)
CO)DC
0.C
a)0
a)
0a
80 _
60 F
40 .
201
0
0 20 40 60 80 100 120
Time after electroporation(minutes)
Figure 3. Effect of uptake of (0) Quin-2 and (0) aequorin on proto-plast plasma membrane resealing after electroporation. D. carotaprotoplasts were electroporated (at 0 min) in the presence of Quin-2or aequonn as described in "Materials and Methods" and then incu-bated for a further 2 h. At intervals during this incubation, samples ofprotoplasts were removed and permeability measured. Plasma mem-brane integrity (permeability) was assessed by the failure of proto-plasts to exclude 0.01% ethidium bromide for 1 min.
trols. These data indicate that after electroporation and indi-cator loading, protoplast plasma membrane integrity was
restored.Quin-2 loading depressed total ATP levels to 30% that of
unloaded controls as shown in Figure 4. EGTA had a similareffect to Quin-2 (Fig 4). Thus, the reduction in ATP contentmay reflect the Ca2' buffering action of intracellular Quin-2,or EGTA. Aequorin was less disruptive, reducing total ATPlevels to 80% of unloaded controls (Fig 4).
Protoplasts loaded with Quin-2 failed to divide when sub-sequently cultured. In contrast, aequorin loaded protoplastswere capable of cell division. Their plating efficiency, at 7 d,was 23.2% compared with 31.6% for unloaded controls.
Indicator Calibration
The response of the Ca2+ indicators to changes in Ca2+ levelcan be calibrated in two ways: (a) with the isolated indicatorin solutions of known Ca2+ concentration-an external cali-bration; (b) with the indicator loaded into the cells and thensetting the intracellular Ca2+ to known levels using ionomycinand Ca2' EGTA-an internal calibration. The internal cali-bration has the advantage of assessing the Ca2" responsivenessof the indicator in the same cellular environment in whichCa2' estimations are to be made.
Figure 5 compares the internal and external calibration ofthe indicator signal, rate of aequorin consumption or Quin-2fluorescence, against Ca2+ concentration. The protocols are
described in "Materials and Methods". The internal andexternal calibrations of Quin-2 differed by less than 10% (Fig.5A). However, the relative magnitude of changes in the rate
Figure 4. Effect of uptake of (0) Quin-2, (@) EGTA, and (A) aequorinon protoplast total ATP levels. Protoplasts of D. carota were electro-porated (at 0 min) with a single 3 kv cm-' D.C. pulse in the presenceof 10 mm Quin-2 or EGTA or 1 mg mLV1 Ca2+-discharged aequorin.ATP was assayed with the luciferin/luciferase reaction at successivetime intervals after the electropermeabilization.
ofaequorin discharge on altering the Ca2+ concentration were
greater if aequorin was calibrated free in solution (externalcalibration) rather than intracellularly (internal calibration)(Fig. SB). The density of the protoplast suspension used forthe calibration of aequorin was 20-fold higher than for Quin-2. Therefore, it is possible that the discrepancy between theinternal and external calibrations of aequorin but not Quin-2was caused by homeostatic buffering of the extracellular cal-cium level by the denser protoplast suspension. Alternatively,a cellular constituent not allowed for (or present at an incor-rect concentration) in the external calibration solution mayhave interfered with the Ca2' dependent light emission ofaequorin (26) causing the response of the intracellular indi-cator to apparently differ from that of the free form. Becauseof the greater certainty of the free [Ca2+] set in the externalcalibration, this method was adopted for all quantitativedeterminations of intracellular [Ca2"] using aequorin. Anexternal calibration has been successfully applied by othersusing aequorin with both animal (5, 24) and algal cells (28).Internal calibrations of aequorin have rarely been reported.
Intracellular Localization of Quin-2
The reliable estimation of cytosolic calcium levels requiresindicators which are retained in the cytosol. Bush and Jones(4) have detected the accumulation of Fura 2, but apparentlynot Indo 1, in vacuoles, and possibly other organelles, ofbarley aleurone protoplasts.The vacuole represented only some 20% of total protoplast
volume of the D. carota protoplasts used in this study. Thiswas estimated from relative areas of vacuoles to cytoplasm inrepresentative electronmicrographs (data not shown). To de-termine if the Quin-2 taken up by the D. carota protoplasts
.0
a)IL.
~a~aII)
pCac
a0
O internalE
0
2CIC0
n. 3xternala)co
0
a4-co
9 8 7 6 5 4 3pCa
Figure 5. Intemal (0) and extemal (0) calibration of rate of aequorinuseage (A) and Quin-2 fluorescence (B) to calcium concentration.Calibration was performed using both Blink's and Sigma aequorin,which gave similar results. The data presented are for Blink's aequorinfor the intemal and Sigma aequorin for extemal calibration. [Ca2+]was set with 5 mm EGTA with CaCI2 to give the appropriate free[Ca2+]. The detailed calibration protocols are described in "Materialsand Methods." The light emission from aequorin has been normalizedfor the variation in the levels of the photoprotein present in differentexperiments by division by L (see "Materials and Methods"). pCais -log10[Ca2].
remained in the cytosol or was localized into organelles, D.carota protoplasts were loaded with [3H]Quin-2 by electro-poration and resealed for 1 h. They were then fractionatedinto cytosol, intact vacuoles and a 'pellet' which containedorganelles. The distribution of radioactivity was then deter-mined. The data are shown in Table I. Of the recovered [3H]Quin-2 70.1% was associated with the cytosolic fraction(though this fraction was contaminated by vacuolar contents,Table I), 8.2% with vacuoles, and 2.8% with the pellet. Thus,while most ofthe Quin-2 remained in the cytosol, a significantproportion was associated with the vacuoles.
Table II shows the results ofa complementary in vitro studyinvestigating the uptake of [3H]Quin-2 by isolated vacuoles.After 60 min incubation, 5.3% of the added [3H]Quin-2 was
either taken up by these isolated vacuoles or became stronglyassociated with them. Vacuolar association was stimulatedsixfold by 100 AiM ATP. This may reflect its active uptake byan anion transporter, possibly energized by ATP hydrolysis.However, ATP may have also stabilized the isolated vacuolarmembranes, leading to an enhanced retention of any Quin-2taken up. Any vacuolar uptake of Quin-2 would superimposevacuolar Quin-2 fluorescence on any cytosolic signal makingcytosolic measurements inaccurate. The limited yield of vac-uoles from Quin-2 loaded protoplasts made direct estimationof their fluorescence signal impracticable.The protein nature of aequorin makes intracellular distri-
bution analogous to that of Quin-2 unlikely. To measure thisdistribution, aequorin was labeled with 125I. However, therelatively small amounts ofthe iodinated protein taken up on
electroporation and the low specific activity of the iodinatedaequorin precluded a fractionation study similar to that re-
ported above for Quin-2. It is generally assumed that once
introduced to the cytosol, aequorin remains there (5, 28);however, this point requires further study-for example, bydirect visualization of aequorin loaded cells with a suitableimage intensification system.
Basal Cytoplasmic Calcium Levels
Table III shows a direct comparison of basal cytoplasmiccalcium levels estimated by Quin-2 and aequorin in proto-
plasts of D. carota and H. vulgare. Barley protoplasts were
more highly vacuolated than carrot (80 and 20% of totalprotoplast volume, respectively) but cytoplasmic calcium,measured with Quin-2, was lower. Our earlier studies on mungbean root protoplasts (which had a vacuolar proportion ofapproximately 50% of protoplast volume) gave a cytoplasmicCa2+ level of 171 + 41 nM, n = 15 (7). So, apparent cyto-plasmic calcium concentration does not correlate with vacu-
olar proportion of the cell. This is perhaps unexpected if thevacuolar uptake of the indicator were to significantly contrib-ute to the cellular Ca2+ signal. The basal [Ca2+] was belowthat accurately measureable with aequorin in our system (lessthan 200 nM) which therefore represents the upper limit forcytosolic Ca>2 levels in the aequorin loaded protoplasts.
Figure 6 shows the variation in measured intracellular levelsof calcium upon varying the extracellular calcium concentra-tion. Intracellular Ca2 was stably maintained over the range
I0- to I0 M extracellular Ca2. However, below 10-7 M,extracellular Ca2 intracellular levels fall. This reduction was
only revealed with Quin-2 loaded protoplasts as, throughoutthe range of extracellular Ca>2 concentrations used, the basalCa> level was below the 200 nm lower limit for aequorin
measurements. At concentrations of 100 mM extracellularCa2, the Quin-2 fluorescence signal considerably increased(data not shown). This may reflect enhanced uptake of cal-cium at this concentration. However, the effects of such highcalcium levels in the medium could also result from thecytotoxic breakdown of cellular control mechanisms.
Azide Increases Intracellular Calcium Levels
We have previously noted (8) that calcium levels sponta-neously increase during the 15 to 20 min of continuousmeasurements with Quin-2 loaded protoplasts. Similarly, dur-ing [Ca"] measurement with aequorin, the basal [Ca2+], wasobserved to increase after 15 to 30 min (Fig. 7). A similar,but more rapid, [Ca>2] increase could be induced on appli-cation of the respiratory inhibitor NaN3 to aequorin loadedprotoplasts. When Quin-2 loaded protoplasts were treatedwith NaN3 Quin-2 fluorescence was quenched. An acidifica-
Table I. Distribution of 3H-Quin-2 in Fractions from D. carota protoplastsProtoplasts were electroporated in the presence of [3H]Quin-2. After 1 h resealing incubation the
protoplasts were lysed and fractionated into supematant, vacuoles, and pelletable organelles. 107protoplasts contained: 372 cpm 3H-Quin-2, 0.04 absorbance units min-' PEP carboxylase activity, 0.01absorbance units min-' fumarase activity, and 142 units of 5-CF fluorescence. These values were takenas 1 00%.
Percent Recovery of
Fraction 3H-Quin2 PEP carboxylase Fumarase 5-CFactivity activity activity fluorescence
Table Ill. Basal [Ca2+J Measured with Aequorin and Quin-2 inBarley and Carrot Protoplasts
[Ca2-] below 200 nm cannot be accurately deternined with ae-quonn. Data represents mean ± SE. The estimated percentage oftotal protoplast volume represented by vacuoles was 20% for D.carota and 80% for H. vulgare.
[Ca2+], Measured byProtoplast System
Quin-2 Aequorin(nM)
H. vulgare 120 ± 62, n = 3 <200, n = 4D. carota 361 ± 47, n = 18 <200, n = 5
0.5%-
E0._
c00._
co
c
00
CU0.co
0.41
0.3 F0.2-
0.11-
08 7 6 5 4 3 2
pCaFigure 6. Effect of extracellular [Ca2+] on intracellular [Ca2+] meas-ured with Quin-2. Extracellular [Ca2+] was set with 5 mm EGTA andCaCI2 to give the appropriate free [Ca2+], calculated by a reiterativecomputer program. pCa is -log1o[Ca2+]. Equivalent experiments withaequorin loaded protoplasts gave basal levels of Ca2+ below the 200nm lower detection limit throughout the range of extracellular calciumconcentrations used.
tion of the cytosol could be responsible since Quin-2 fluores-cence is sensitive to low pH. Bush and Jones (4) have notedboth cytoplasmic acidification and quenching of Indo 1, aCa2+-indicator related to Quin-2, in barley aleurone proto-plasts. Azide had no measureable effect in vitro on Quin-2/Ca2+ fluorescence or aequorin luminescence. Thus, the spon-
00
0C.)
E0GoC
0
0
coC
0
0C000co0)0
0
1
21
4
5
0 5 10 15 20Time (minutes)
25
210.0 %
E1.0 .5
0.5 '0E0
0.90.2 26-C.
Figure 7. Effect of 1 mm NaN3 (added at arrow) on intracellular Ca2+levels of aequorin loaded protoplasts of D. carota, indicated by therate of discharge of aequorin. The control was aequorin loaded butnot treated with azide. Calcium concentrations below 200 nm cannotbe accurately estimated with aequonn.
taneous increase in Ca24 levels during prolonged incubationnoted above may have arisen from the onset of anoxia in thedense protoplast suspension required for Ca24 measurement.
DISCUSSION
Ca2+ has been implicated in the regulation of numerousphysiological processes and enzyme activities in plants (6, 12).However, technical problems with directly measuring cyto-plasmic Ca2" levels have hampered investigations as to theprecise role of this ion in cellular regulation. We have com-pared two indicators of intracellular Ca2+ levels, aequorin andQuin-2, in higher plant cell protoplasts. Both indicators hadadvantages and disadvantages. Quin-2 was observed to depressD. carota protoplast ATP content by 60 to 70% (as doesEGTA) and interfered with subsequent cell division, whereasaequorin reduced ATP by only 20% and did not inhibitdivision. The parallel effects of Quin-2 and EGTA suggestthat the disruptive effects of Quin-2 may arise from its Ca2+-buffering action. Cytoplasmic calcium control systems couldbe modified, affecting, for example, the Ca24 regulated stepsin mitochondrial ATP synthesis (18). A similar depression inATP levels has been observed in animal cells loaded withQuin-2 (24). The Kd ofaequorin for Ca24 is I0 M. In contrast,the Kd for Quin-2 and EGTA for Ca24 is 10' M (26, 27). Also,the intracellular concentration of aequorin, probably lessthan 10-7 M (assuming uptake of 1% of total protoplastvolume (0.5 pL) on electroporation), was much lower thanthe 10' M intracellular Quin-2 levels required for measure-ment of intracellular Ca2` concentration (7, 8). Therefore, thereduced effect of intracellular aequorin on protoplast func-tion may reflect its reduced Ca24-buffering capacity comparedto Quin-2.
Thus, the less disruptive nature of intracellular aequorinwould seem to make it preferable to Quin-2 as a potentialindicator of increases in cytoplasmic Ca2+ levels. Despite this,resting levels of calcium appeared to be low and stable inQuin-2 loaded protoplasts over a wide range of extracellularcalcium levels, indicating that cellular Ca2" regulation wasstill being rigorously maintained.The vacuolar association of a fraction of the loaded Quin-
2 (typically 10% in D. carota protoplasts) represents anotherputative disadvantage to the use of this indicator. Fluores-cence from vacuolar Quin-2 would make quantitative analysisof the Ca2' dependent fluorescence signal from loaded cellscomplex and not a direct measure of [Ca2J]i. We were unableto isolate sufficient vacuoles to measure any associated Quin-2 fluorescence. However, vacuolar fluorescence may not sig-nificantly contribute to the measured Quin-2 signal. Vacuolescontain heavy metals and maintain a low internal pH. Bothof which could reduce vacuolar Quin-2 fluorescence (5). Thecells used in this study were grown in 100 AM Mn2+ ions. Ifthe vacuole reached an equivalent ion concentration its Quin-2 fluorescence would be completely quenched.The Kd of Quin-2 for Ca2" is I0-` M. Therefore, at the
presumed vacuolar concentration of Ca2", 1 mM (12), anyvacuolar Quin-2 would be completely saturated and maxi-mally fluorescent (Fig. 5). A change in vacuolar calciumwould not register as an altered fluorescence signal until theconcentration declined to below 10 ,uM. Thus, when Quin-2fluorescence increases, as for example in Figure 1, this alteredsignal is unlikely to originate from the vacuole.Aequorin reports a similar elevation in calcium caused by
ionomycin treatment. The major signal then from both Quin-2 and aequorin probably originates from indicator in a com-partment at a [Ca2] much lower than the extracellular con-centration, such as the cytosol. We have previously reported(8) increases in Quin-2 fluorescence with various calmodulinbinding inhibitors. Again it would be difficult to see how thiscould originate if the Ca2+/Quin-2 signal was simply vacuolar(calmodulin binding inhibitors inhibited aequorin lumines-cence so that confirmation of these Quin-2 data was notpossible).
Similarly, on application of ionomycin to Quin-2-loadedprotoplasts, the fluorescence signal rose rapidly from a stableresting level to an increased but stable level (Fig. 1B). Thereis no evidence of multiple phases to this increase as might beexpected if several cellular compartments were affected by thetreatment. This suggests most of the fluorescence signal arisesfrom Quin-2 in a single, rapidly equilibrating, cellular frac-tion, most probably the cytosol. We have, therefore, con-cluded that the major fluorescence signal from Quin-2 loadedprotoplasts is from the cytosol. Any Quin-2 signal from thevacuole may simply overestimate the resting level. However,our data indicate that the more highly vacuolated protoplastsdo not have higher basal calcium levels. The Quin-2 familyof Ca" indicators may be more suited to measurement sys-tems where any intracellular localization can be observed andexcluded from the cytosolic [Ca"] measurements, such aswith an image analysis system. Indeed, Fura-2 seems to crossorganelle membranes in both plant (3, 4) and animal cells(24). It seems unlikely that aequorin, a 20 kD polypeptide,
would exhibit a similar intracellular distribution to Quin-2.This point requires further investigation; for example, bydirect visualization of light emission from loaded cells with asuitable image intensification system.
It can be seen from the calibration of aequorin light emis-sion against calcium level (Fig. 5A) that aequorin cannot beused to accurately measure calcium levels below 2 x 10' M.Thus, in our measurement system, aequorin seems best suitedto quantify increases in the basal cytoplasmic calcium levelsuch as those induced by azide (Fig. 7). A similar restrictionin only reliably quantifying increases in cytoplasmic Ca2 isalso shown by the metallochromic [Ca2+], indicators (26).Nevertheless, the metallochromic indicator arsenazo III hasproved extremely useful in revealing transient increases in[Ca2+]i during mitosis in Tradescantia stamen hairs (I 1) andaequorin has been successfully used to measure transientincreases in [Ca2+]i in algal cells (22, 28). To determine basalcalcium levels and relatively small changes in level, as havebeen reported, for example, on illumination of Nitellopsiscells (20) the Quin-2 family of indicators (despite their draw-backs) or, where feasible, Ca2+-selective microelectrodes maybe preferable.
Cytoplasmic Ca2 levels were observed to increase on theaddition of azide and potentially on the development ofanoxia in the sample cuvette, (Fig. 7). This may reflect therelaxation of the rigorous maintenance of the low basal Ca>2as cellular energy is depleted. Alternatively, gradual protoplastdeath may have given rise to a small fraction of cells with anextremely high intracellular Ca>2 level. An increase in cyto-plasmic [Ca>], potentially to cytotoxic levels (greater than10-6 M, ref. 12) could help to explain some of the effects of,for example, azide in the breakage of seed dormancy (25) oranoxia and flooding on plant development and viability (1).However, azide treatment caused the quenching of intracel-lular Quin-2 fluorescence; similar quenching of Indo 1 fluo-rescence has been reported (4). Therefore, cross-checking ofthe results obtained with aequorin using Quin-2 was notpossible.Aequorin and Quin-2, and its related indicators, may be
seen as complementary methods of measuring [Ca2+],. Quin-2 is capable of the indicating low basal cellular Ca>2 levelswhereas aequonn cannot quantify levels below 10-7 M. How-ever, aequorin appears less disruptive of cellular function andmay be applicable when Quin-2 is not, such as with azideeffects, and vice versa. Thus, these indicators have bothadvantageous and disadvantageous characteristics, dependingon the application.Our data also directly indicate that ionomycin and Ca-
EGTA can modulate cytoplasmic calcium levels. This shouldprove valuable in attempting to discriminate the calciumdependency of the initiation of metabolic changes in vivo. Wehave also shown the uptake of a protein through electroporesin the protoplast plasmamembrane. This opens the possibilityof adjusting the protein composition of the higher plantcytosol with enzymes and other proteins.
ACKNOWLEDGMENT
We would like to thank A. Mitchell for the preparation of barleyprotoplasts.
1. Armstrong W (1979) Aeration in higher plants. Advances in BotRes 7: 225-332
2. Brownlee C, Wood J (1986) A gradient of cytoplasmic freecalcium in growing rhizoid cells ofFucus serratus. Nature 320:624-626
3. Brownlee C, Wood JW, Briton D (1987) Cytoplasmic free cal-cium in single cells of centric diatoms: the use of Fura 2.Protoplasma 140: 118-122
4. Bush DS, Jones RL (1988) Measurement ofcytoplasmic calciumin aleurone protoplasts using Indo- I and Fura-2. Cell Calcium8: 455-472
5. Cobbold PH, Rink TJ (1987) Fluorescence and bioluminescencemeasurement of cytoplasmic free calcium. Biochem J 248:313-328
6. Gilroy S, Blowers DP, Trewavas AJ (1987) Calcium: a regulationsystem emerges in plants. Development 100: 181-184
7. Gilroy S, Hughes WA, Trewavas AJ (1986) The measurementof intracellular calcium levels in protoplasts from higher plantcells. FEBS Lett 199: 217-221
8. Gilroy S, Hughes WA, Trewavas AJ (1987) Calmodulin antag-onists increase free cytosolic calcium levels in plant protoplastsin vivo. FEBS Lett 212: 133-137
9. Gould GW, Coyler J, East JM, Lee AG (1987) Silver ions triggerCa24 release by interaction with the (Ca24-Mg2i-ATPase inreconstituted systems. J Biol Chem 262: 7676-7679
10. Hall JL, Moore AL, eds (1983) Isolation of Membranes andOrganelles from Plant Cells. Academic Press, New York
11. Hepler PK, Callaham DK (1987) Free calcium increases duringanaphase in stamen hair cells of Tradescantia. J Cell Biol 105:2137-2143
12. Hepler PK, Wayne RO (1985) Calcium and plant development.Annu Rev Plant Physiol 36: 397-439
13. Kanchanapoom K, Boss WF (1986) The effect of fluorescentlabelling on calcium-induced fusion of fusogenic protoplasts.Plant Cell Rep 5: 252-255
14. Keith CH, Raten R, Maxfield FR, Bajer A, Shelanski ML (1985)Local cytoplasmic calcium gradients in living mitotic cells.Nature 316: 848-850
15. Kringstad R, Kenyon WH, Black CC (1980) The rapid isolationofvacuoles from leaves ofcrassulacean acid metabolism plants.Plant Physiol 66: 379-362
16. Lieberman M (1979) Biosynthesis and action of ethylene. AnnuRev Plant Physiol 30: 533-591
17. Liu C-M, Hermann TE (1978) Characterization of ionomycin asa calcium ionophore. J Biol Chem 253: 5892-5894
18. McCormack JG, Denton RM (1986) Calcium as a second mes-senger in mitochondria. Trends Biochem Sci 11: 258-262
19. Mehrle W, Zimmermann U, Hampp R (1985) Evidence forassymmetrical uptake of fluorescent dyes through electro-per-meabilised membranes ofAvena mesophyll protoplasts. FEBSLett 185: 89-94
20. Miller AJ, Sanders D (1987) Depletion of cytosolic free calciuminduced by photosynthesis. Nature 326: 397-400
21. Nobiling R, Reiss HD (1987) Qualitative analysis of calciumgradients and activity in growing pollen tubes of Lilium lon-giflorum. Protoplasma 139: 20-24
22. Okazaki Y, Yoshimuto Y, Hiramuto Y, TazawaM (1987) Turgorregulation and cytoplasmic free Ca2+ in the alga Lamprotham-nium. Protoplasma 140: 67-72
23. Owen JH, Hetherington AM, Wellburn AR (1987) Calcium,calmodulin and the control of respiration in protoplasts iso-lated from meristematic tissues by abscisic acid. J Exp Bot 38:1356-1361
24. Pozzan T, Rink TJ (1985) Using Quin-2 in cell suspensions. CellCalcium 6: 133-144
25. Roberts EH (1972) Oxidative processes and the control of seedgermination. In W Hetdecker, ed, Seed Ecology, Butterworths,London, pp 189-213
26. Thomas MV (1982) Techniques in Calcium Research. AcademicPress, London
